IV Key Issues in Official Safety Nets and International Financial Markets
- International Monetary Fund
- Published Date:
- January 1991
Throughout the 1970s and 1980s, major domestic and offshore financial markets played an increasingly important role in the financing of large current account and fiscal imbalances in the industrial countries, the allocation of global savings across countries and regions, and the provision of borrowed international reserves. The stability of international capital flows, exchange rate arrangements, and the reserve system has thus become more dependent on the stability of international financial markets and their attendant payments, clearance, and settlement systems.38 In consequence, an extended disruption of these systems could clearly have highly adverse systemic effects on global trade and capital flows.
The stability of national financial systems has traditionally been underpinned by official safety nets designed to prevent financial disturbances from spilling over onto the real economy. While the nature of these safety nets has differed across major industrial countries, they have generally encompassed the provision of short-term emergency liquidity assistance by central banks, some form of private or official deposit insurance (to reduce the incentives of depositors to rapidly withdraw funds during a crisis), and direct short- or medium-term emergency assistance for large troubled financial institutions. While such policies can help contain the effects of a financial crisis, they expose the authorities to credit risks either through lending at the central bank’s discount window, lending to troubled financial institutions by the government, or deposit insurance obligations. To limit the authorities’ exposure to such credit risk, official safety nets have also incorporated policies specifying minimum capital adequacy standards for financial institutions, systems of prudential supervision, and limits on risk taking by institutions and individuals.
The increased integration of financial markets, both within and across countries, has raised new questions about the design and cost-effectiveness of national safety nets. As financial liberalizations in the major industrial countries have removed restrictions on domestic and cross-border financial transactions, there has been a sharp expansion in international capital flows, the domestic activities of foreign financial entities, the degree of competition in domestic financial markets, and the pace of financial innovation. As a result, concern has been heightened about the credit risks to which governments are exposed through official safety nets. Specific issues are (1) whether existing systems of prudential supervision can adequately monitor the activities of domestic institutions in foreign markets and foreign firms in domestic markets; (2) whether foreign institutions should be allowed (or required) to participate in national deposit insurance arrangements; (3) whether deposits in the foreign branches and subsidiaries of domestic financial institutions should be protected by the domestic deposit insurance system; (4) which central banks, if any, would be willing to provide emergency liquidity assistance in offshore markets; (5) whether official deposit insurance and other guarantees might stimulate greater risk taking by some financial institutions or individuals; and (6) what steps, in general, might be needed to ensure the cost-effectiveness of official safety nets.
This section examines some of these issues by first considering the rationale for official safety nets and then reviewing the key components of the safety nets in the major industrial countries. The effects of these safety nets on financial stability and risk taking in the private sector are next discussed, and finally, there is an examination of recent reform proposals that have been made in the United States to improve the cost-effectiveness of its official safety net.
Rationale for Official Safety Nets
During the twentieth century, most major industrial countries experienced periods of significant financial instability, particularly in the 1930s and in periods surrounding major wars, that involved widespread failures of financial institutions, sharp declines in asset prices, and disruptions of both national payments, clearance, and settlement systems and the intermediation of credit. The perception that such financial instability contributed to sharp declines in real activity and employment led the authorities in the major industrial countries to establish policies and institutions that created an official safety net designed to promote financial stability and to limit the spillover effects of financial crises onto the real economy.
Much of the consequent debate on the need for official safety nets has focused on the contagion or systemic risk that the failure of illiquid or legally insolvent financial institutions can create. An example of contagion is a banking panic in which there is a sudden increase in the perceived riskiness of holding deposits, and depositors demand a large-scale conversion of deposits into currency. In the absence of an official safety net, adverse information about the value of an institution’s portfolio, such as news of the bankruptcy of a principal borrower, could lead forward-looking depositors to participate in a run to avoid trying to withdraw their funds after the institution’s liquid assets have been exhausted. Similarly, widespread banking panics could occur if many borrowers fail in a severe depression or if the failure of one institution raises concerns among depositors about the viability of other institutions because depositors are imperfectly informed about the viability of individual banks.39
Avoiding the real costs of a banking (or liquidity) crisis has been an important consideration in formulating public policies. Although the exact magnitude of these costs is subject to dispute, the perception of policymakers is that they can be substantial. This reflects the view that it is costly, if not impossible, for participants in the interbank market and other wholesale and retail depositors to distinguish between an illiquid but solvent bank and an economically insolvent institution. As a result, banks, in general, and the interbank market, in particular, are unlikely to be able to organize a means of allocating liquidity in the event of a crisis.40
An alternative view is that banking crises are rare events and that an economically solvent bank generally experiences little difficulty in obtaining sufficient liquidity by borrowing from other banks or possibly selling assets to meet its maturing liabilities.41 Moreover, in the event of economic insolvency, a bank run ensures the timely closure of a troubled institution. This perspective emphasizes the role of the interbank market for high-powered money and its ability to perform well in periods of increased uncertainty about the riskiness of banks. Provided that there is not a widespread increase in the demand for currency by the nonbank private sector, the interbank market in principle can allocate high-powered money to meet the liquidity needs of individual banks, using collateralized loans if necessary. A central bank serving as lender of last resort would then be required to provide liquidity only in the unlikely event of a failure of the interbank market or a widespread portfolio shift from bank deposits to currency.
While banking panics provide highly visible, historical examples of systemic risk, much of the current concern about contagion focuses on national and international payments, clearance, and settlements systems. Explicit and implicit deposit insurance guarantees in many countries have significantly reduced the incentive for retail depositors to participate in a bank run. As the scale of global securities and foreign exchange trading has grown rapidly in the past two decades, the credit risks inherent in the operations of these payments, settlements, and clearance systems have increased accordingly, creating concern that there is now a greater potential for liquidity crises and contagion. In most payments systems, for example, the potential for contagion exists because any party with access to the system may not have sufficient funds to meet its payments obligations at time of settlement,42 and such a counterparty failure could create payments difficulties for a broad range of other participants.
In addition to reducing the costs associated with liquidity crises and disruptions to payments, clearance, and settlement systems, a broader rationale for an official safety net is the view that the authorities may have to act to preserve confidence in the financial system, sustain the credit intermediation process, and prevent extreme losses of wealth in the private sector.43 This view envisions an official safety net that extends well beyond provision of temporary liquidity to money markets and could potentially encompass government assistance for important nonbank as well as banking institutions and possibly support of asset prices (including equity prices)44 in the event of a financial crisis.
Proponents of this view have argued that assistance for important nonbank financial institutions (such as large securities houses) may at times be needed to prevent a major disruption of the financial intermediation process, in general, and the extension of credit to businesses, in particular. While the failure of a single nonbank financial institution might affect only a narrow set of borrowers, it has been suggested that widespread failures of such institutions would disrupt normal creditor-borrower relationships and, in the face of the general uncertainties created by a crisis period, borrowers would find it difficult to find new sources of funds. Such a credit “squeeze” would force nonfinancial firms to sharply curtail their output and employment.
The view that, on some rare occasions, the authorities may need to intervene in asset markets generally reflects concern that large, abrupt declines in asset prices can (1) have a highly deflationary effect on consumer spending (through wealth effects) and investment activity (because of increased uncertainty)45 and (2) create widespread bankruptcies of financial institutions (including banks) that are holding these assets. Although it has been recognized that the authorities cannot stand against the tide if economic fundamentals imply that asset prices must fall, some analysts have argued that there are times when asset markets become “disorderly” and bid-ask spreads widen drastically in response to imperfect information. In these situations, timely official assistance may help to maintain orderly markets and prevent unwarranted asset price volatility. In this regard, it is often noted that governments and central banks in the major industrial countries regularly intervene in government securities and foreign exchange markets by acting as a willing buyer or seller of last resort when these markets are viewed as disorderly.
It is not clear what influence these narrow and broad views on the role of official safety nets have had on the design and evolution of official safety nets in the major industrial countries. To an important degree, this reflects the fact that the authorities have often been purposely vague about the conditions under which some types of official assistance will be available. This vagueness arises because of concerns that the certainty of official assistance could induce a pattern of risk taking in the private sector that would ultimately require official assistance to financial institutions or markets that could be expensive for taxpayers. Indeed, a fundamental dilemma is that, while official assistance during a financial crisis can limit the effects of the crisis on real activity and income, the knowledge or expectation that such assistance will be available may alter the behavior of managers, creditors, depositors, and owners of financial institutions in such a way as to make a crisis more likely, which could create large future credit risks for the authorities. While the dilemma exists in all financial systems, the diversity of the institutional, supervisory, and legal arrangements that comprise official safety nets in the major industrial countries suggests that the authorities in these countries have followed somewhat different paths to confronting this dilemma.
Key Components of Official Safety Nets in Major Financial Systems
Official safety nets in the major industrial countries have encompassed (1) the provision of emergency liquidity assistance and official guarantees to the financial system to promote financial stability and (2) systems of prudential supervision, regulation, and capital adequacy requirements to limit the authorities’ exposure to credit risks.
Sources of Official Assistance and Guarantees
Managing liquidity crises has involved the central bank’s traditional function of providing emergency liquidity assistance to the money markets and the provision of assistance to specific institutions whose failure could have widespread systemic consequences. Since a liquidity crisis can produce sharp increases in holdings of currency and interest rates, central banks have traditionally stood ready to inject short-term liquidity into the market through open market purchases of securities or through the acceptance of securities at the discount window from individual institutions affected by depositor withdrawals.46 In general, access to the discount window has been limited to licensed deposit-taking institutions.
In addition to providing emergency assistance during a general liquidity crisis, the authorities have at times provided special assistance to large institutions even when there was not a widespread crisis. Such assistance can take many forms, including enlarged access to the central bank’s discount window, forming a support group of banks to extend special loans, helping to arrange the merger of the weak institution with a stronger institution, providing an infusion of capital into weak institutions, and government takeover of the institution.
There has been considerable debate over what types of institutions, if any, should be included in this “too-large-to-fail” category and what effects this policy has on private sector behavior. While large banks have typically received some form of private and official assistance during times of difficulty,47 the provision of such assistance for small and medium-sized banks and even large securities houses (e.g., Drexel Burnham Lambert) has been less certain.48 A key problem lies in identifying the circumstances under which an institutional failure would create a systemic problem. This is likely to depend on the role of the institution in national and international payments systems, whether it acts as a significant component of the clearance and settlement systems in a particular market (e.g., as a settlement bank in a futures or options market), and whether its failure would seriously undermine public confidence in the stability of comparable institutions. The authorities in all major industrial countries have been purposely vague about when (if ever) assistance would be made available to large institutions because of concerns that such assistance could weaken the incentives for depositors, creditors, and the owners of large financial institutions to monitor the activities of the managers of these institutions.49
Moreover, assistance to large institutions has seldom provided full protection for all creditors, depositors, managers, and owners. Although practices have differed across the major industrial countries, institutions that have required official assistance have typically had their senior management replaced, and owners or subordinated debtor bond holders have been assessed all or some significant proportion of the losses incurred by the institutions. Small depositors have usually been fully protected, while the position of large depositors and other creditors has been less clear. Although most deposit insurance systems limit the size of the deposit that is protected, large depositors have often been protected when large banks (e.g., Continental Illinois, Bank of New England) have experienced difficulty. In contrast, creditors have at times been protected and at other times have absorbed losses, especially if the creditors are holders of subordinated loans or bonds (which are counted as part of the institution’s capital in most industrial countries). When the precedent has been established that creditors will be forced to absorb losses, then institutions that have weak capital and income positions have found it expensive to raise capital through subordinated debt. In the United States, for example, the cost of servicing subordinated notes issued by some money center banks, whose credit rating had been reduced, has recently risen to about 400–500 basis points over the interest rate on comparable maturity U.S. Treasury securities.
While the central bank’s discount window has often been viewed as an effective means of meeting a short-run liquidity crisis, all major industrial countries now have some form of deposit insurance to provide depositors with at least partial protection against bank insolvency.50 Although deposit insurance was introduced in the United States in 1933, most European countries and Japan did not have explicit deposit insurance programs until after the mid-1960s. For example, the Federal Association of German Commercial Banks established the Deposit Protection Fund for private commercial banks in 1966, and the French Bankers Association created a private insurance scheme in 1980. These contrasting experiences reflected a number of historical factors. The United States had introduced deposit insurance following a series of banking panics that had resulted in widespread bank failures and a general loss of confidence in the banking system. Since the U.S. financial system included a large number of small and medium-sized banks, it was felt that general confidence could be restored to the system as a whole only if depositors were protected regardless of the size of the bank in which they placed their funds.
The absence of explicit deposit insurance in Europe and Japan during the 1950s and 1960s did not mean that depositors, especially small depositors, perceived that they were exposed to large risks. In that period, regulatory restrictions on the activities, location, instruments and, in some countries, interest rates of financial institutions limited both competition and portfolio choices of financial institutions. Restrictions on competition enabled protected institutions to strengthen their income and capital positions; whereas restrictions on activities and instruments were viewed as a means of limiting risk taking.51 Tight prudential supervision, especially of large financial institutions, featured prominently in Japanese and European systems. Moreover, banks that got into difficulty were typically assisted directly or indirectly (e.g., through merger) so that at least small depositors did not suffer any losses.52 The move toward explicit deposit insurance (whether private or official) in Europe and Japan appears to have reflected in part the desire of depositors for more certain protection and of the authorities for a means of financing heretofore implicit deposit guarantees in a world in which restrictions on the domestic and foreign activities of financial institutions have been relaxed and in which there have been periods of asset price volatility and global financial shocks.
The structures of the deposit insurance systems in the major industrial countries nonetheless differ widely in terms of the amount of insurance provided, the institutions allowed or obligated to participate in the systems, the relative roles of private and official sources of insurance, and the extent to which insurance funds have been utilized.53 The most important similarity between the deposit insurance systems is that they all have a formal limit stipulating that only deposits up to a certain size are insured.54 Moreover, while nonbank, nonresidents’ deposits in domestic banks are covered (up to the formal limit) by the deposit insurance systems in all major industrial countries, interbank deposits are not formally covered by the bank insurance funds in France, Germany, Japan, and the United Kingdom. In addition, although all five of the largest industrial countries allow or require the deposit-taking domestic branches of foreign banks to participate in their insurance systems, some have not formally extended coverage to deposits in foreign branches of domestic banks. Finally, while some countries (such as the United States) rely solely on official insurance systems, other countries (such as Germany and France) rely on banking association arrangements.
While the presence of a lender of last resort and deposit insurance have played a key role in preventing banking panics, their existence has forced the authorities to assume some credit risk and created concerns about a potential “moral hazard” problem. As with other types of insurance, an official safety net must confront the problem that the insured may behave differently just because of the existence of insurance. If fully insured, depositors would have little reason to worry about or to monitor the riskiness of their bank’s activities.55 In this situation, depositors would not demand that a risk premium be incorporated into the yield they receive on their deposits as the riskiness of a bank’s portfolio increases, and an element of market discipline on banks would thus be weakened. As a result, the cost of using deposits as a source of funding for lending activities would be reduced relative to the cost of using equity.56 Even in the absence of greater risk taking by financial institutions, increased reliance on deposits would reduce the “buffer” between the cost of an institution’s failure and the obligations of the taxpayer. Moreover, managers of some financial institutions, especially those close to insolvency, might have an incentive to undertake an unduly high share of potentially high return but also high-risk activities under the assumption that, with good outcomes, they will reestablish the financial position of the institution, but, with bad outcomes, the losses will be absorbed by the deposit insurance fund or ultimately by the taxpayer. The effects of this moral hazard problem on risk taking and institutional structures are discussed later in this section.
Measures to Limit Official Credit Risks
Concern about the potential costs of the credit risks associated with official safety nets has led to the implementation of policies designed to limit those risks. The measures have involved the establishment of minimum capital adequacy standards, financial codes, restrictions on risk taking, systems of prudential supervision, and systems for managing the risks inherent in payments, clearance, and settlement systems.
Capital adequacy requirements have been seen as a means of both strengthening the ability of financial institutions to withstand the effects of financial and real sector shocks and helping to overcome the moral hazard problems created by the existence of an official safety net. There is clear historical evidence that, in the absence of official safety nets, financial institutions have held high ratios of capital to assets. For example, in the United States, commercial banks had an average capital/asset ratio that exceeded 20 percent throughout the period from 1844 to 1900.57 Between 1914 (when the Federal Reserve was established) and 1934 (when the Federal Deposit Insurance Corporation (FDIC) was formed), the average capital/asset ratio of banks declined to about 15 percent. By the mid-1980s, this average ratio stood at 6 to 7 percent before beginning to rise again (Table A17). To strengthen the capital positions of banks, the Group of Ten countries in July 1988 reached agreement on new capital adequacy standards for international banks, which will come fully into effect at the end of 1992. The new standards specify the minimum amount of bank capital in relation to the credit risks that banks incur in their on- and most off-balance sheet activities (see Section III).58 Such capital adequacy standards could play a role in improving the cost-effectiveness of an official safety net both by creating an additional buffer between the costs of institutional default and the taxpayer and by introducing an additional element of market discipline by forcing banks to raise capital in the equity markets.
To make the official safety net cost-effective and to ensure that financial markets are perceived to operate fairly, major financial systems incorporate codes of behavior, limits on risk taking by individuals and financial institutions, and prudential supervision arrangements. Financial and commercial codes define broadly the range of activities in which intermediaries may legally engage. An important element in these codes is the establishment of appropriate accounting standards and disclosure requirements. Limits on risk taking have also played a role in efforts to limit the credit risks associated with an official safety net. While the experience of the 1930s led many major industrial countries to establish a variety of restrictions on risk taking in order to create more stable financial systems, the financial liberalizations of the 1970s and 1980s encompassed a weakening or removal of many of these restrictions.59 As discussed in previous International Capital Markets reports, the limitations on risk taking included restrictions on external financial transactions, a separation of commercial and investment banking, segregation of insurance and banking activities, limitations on commercial ownership of banks, prohibitions on the use of certain types of financial instruments (e.g., options), exclusion of individuals and institutions from certain markets, and restrictions on portfolio holdings of various individuals and institutions.60 When combined with appropriate macroeconomic policies and strong prudential supervision, these limitations on activities and portfolio choices were initially perceived as a means of promoting a stable financial system by creating financial institutions with strong market and capital positions, limiting speculative behavior and risk taking, and restricting competition.
By the early 1970s, it was increasingly evident that those restrictions that limited competition were having an adverse effect on the efficient provision of financial services and were distorting the allocation of credit. In addition, exchange restrictions and limitations on holdings of foreign assets were actually increasing the risks confronting domestic residents by preventing them from holding internationally diversified portfolios. As a result, many restrictions on activities, instruments, and portfolio holdings were removed, and increased emphasis was placed on ensuring that market participants have adequate information upon which to base their financial decisions. This process has involved the removal of exchange controls, the relaxation of restrictions on the entry of foreign financial firms into domestic markets, the removal of interest rate ceilings, the introduction of new types of financial instruments (e.g., financial futures and options), and the relaxation of restrictions on competition between different segments of the domestic financial system.61 Some restrictions on risk taking, particularly by individuals, remain in all countries, however, to ensure adequate consumer protection.
Prudential supervision to enforce financial codes and capital adequacy requirements has traditionally been viewed as a necessary complement to lender-of-last-resort assistance in order to make the official safety net cost-effective. Since one of the principal objectives of prudential supervision is to ensure that difficulties confronting a financial institution are corrected before the institution fails, the ability of the regulatory authorities to identify the difficulties and to take prompt corrective action depends on the source of the difficulty, the extent of the deterioration in the institution’s financial position, and the speed with which the problem arises. Some observers have expressed concern that the growing complexity of financial activities and instruments, the increased frequency of global financial shocks, and the expanding scale of foreign and offshore activities of domestic financial institutions have added new obstacles to implementing effective prudential supervision. For example, new financial activities and the use of new financial instruments may make it difficult for the supervisory authorities to gauge the nature of the future risks that are likely to confront the financial institution. Moreover, large macroeconomic shocks may suddenly and adversely affect the financial positions of a broad range of financial institutions. While diversification of a financial institution’s operations across countries can help reduce the effects of shocks occurring in a single country, global or systemic shocks could make it difficult to either implement a gradual program of corrective action or prevent institutional failures.
As the scale of activities of the foreign branches and subsidiaries of domestic financial institutions expanded with the removal of exchange controls and relaxation of restrictions on the entry of foreign firms into major markets, supervisory authorities were also confronted with the task of deciding who had responsibility for supervision of institutions in different markets.62 The dilemma was partly addressed through a multilateral agreement (the Basle Concordat in 1975) between supervisors in the major industrial countries that established supervision of capital adequacy on the basis of consolidated balance sheets for banks and their foreign branches and majority-owned subsidiaries. Other subsidiaries would be supervised by the authorities of the country in which they were located. Collaborative efforts on such matters continue.
An important reason why the financial crises of the 1970s and 1980s in the major industrial countries had only limited short-run effects on real activity is that they did not extensively disrupt major national and international payments, clearance, and settlement systems. Financial liberalization and the growing integration of major markets have led to a sharp increase in the volume of transactions both within and across these systems; consequently, concern has been raised about the ability of existing institutional arrangements to cope with the new volume of transactions and to manage effectively the risks created by counterparty failure and liquidity crises.
At the core of current international monetary arrangements lies an interlocking network of national and international payments systems, which facilitate the exchange of funds associated with almost all international trade and financial transactions. While some domestic payments systems are operated by the central bank (such as Fedwire in the United States), other clearing house systems are operated by private corporations or syndicates. The largest private payments system is the Clearing House Interbank Payments System (CHIPS), which is owned by the New York Clearing House Association and handles U.S. dollar payments associated with foreign exchange transactions involving the U.S. dollar and Eurodollar transfers. In some systems, official and private payments systems are parallel. In Japan, for example, there is both the Zengin system and the Bank of Japan Net system; the Bank of Japan is the settlement bank in both systems.63
In a number of countries, the central bank’s role in the payments system is regarded as an important element in the safety net, especially when the central bank provides for payments finality. Such payments finality allows banks and their customers to make transfers of funds during the day without necessarily having first to immediately provide funds to cover the transfer and to receive funds with the assurance that such transfers will not subsequently be revoked by the failure of the sending party. In operating a payments system that provides such assurances, the central bank typically extends daylight overdrafts to some participants and thereby assumes some credit risk. As the volume of financial transactions expanded with the growing integration of financial markets, the scale of these daylight overdrafts began to grow; authorities in a number of countries have taken steps to limit the credit risk they face.
An example of these risk-reduction efforts is the steps taken by the U.S. Federal Reserve. The Federal Reserve introduced a payments system risk-reduction scheme in 1986 that focused on controlling the direct Federal Reserve credit risk associated with the extension of intraday credit on Fedwire, which had reached a daily average of $65 billion, by establishing a maximum amount of intraday overdraft that depository institutions were permitted to incur.64 In July 1987, the Federal Reserve adopted a two-step 25 percent reduction in these caps. The Board of Governors Large Dollar Payment System Advisory Group then recommended that a fee be imposed on daylight overdrafts65 so as to induce the creation of a private sector market to replace much of the Federal Reserve funding of intraday credit. It was anticipated that such a fee would help reduce the use of daylight overdrafts by delaying less critical payments, shift payments from Fedwire to CHIPS, stimulate greater use of netting arrangements, and further the development of an intraday federal funds market.
While clearing and settlement systems for securities and derivative (options and futures) instruments are generally operated by private clearing houses, these entities may extend significant amounts of credit, especially on an intraday basis. Indeed, the possibility of default by major participants in the settlement systems for stocks and stock index options and futures was regarded as the greatest threat to the financial system during the October 1987 global equity market crash.66 Concern about the stability of these systems led the Group of Thirty, a private group concerned with the working of the international financial system, to propose steps for establishing central securities depositories by 1992, introducing netting systems if they would reduce risk and promote efficiency, and the adoption of a universal settlement date on T + 3 (i.e., three days after the trade date).
Market participants and the Group of Ten central banks have also been studying the feasibility of multilateral netting arrangements for managing better the risks in foreign exchange markets.67 Dealers in foreign exchange typically enter into successive contracts to buy or sell various currencies, often with the same counterparty and the same date of delivery.68 Given the time differences between the hours of operation of major foreign exchange markets, and the chance that one counterparty might fail before settlement occurs, there can be extended periods involving settlement risk.69 In consequence, it has been proposed that multilateral netting be achieved through the creation of a central counterparty or clearing house, whose legal structure would be similar to that of clearing houses for futures and options, and which would help reduce these settlement risks.70
Economic Effects of Official Safety Nets
Although the scope and coverage of official safety nets in the major industrial countries can only be vaguely and imprecisely defined, the perception, often based on historical experience, that some form of official assistance would be available in a financial crisis has influenced the stability of global financial markets, the incentives for risk taking in the private sector, the allocation of credit both within and across countries, and the evolution of institutional structures.
While some thirty separate episodes of system-wide bank runs occurred in France, Germany, Japan, the United Kingdom, and the United States between 1790 and 1929,71 there have not been any system-wide banking panics involving a general flight from bank deposits to cash or to foreign exchange in any major industrial country since World War II. Moreover, major domestic and international payments, clearance, and settlement systems have not experienced extended disruptions. Nonetheless, authorities in the major industrial countries have had to confront a number of financial crises during the 1970s and 1980s that involved failures of both large and small financial institutions.72 While each crisis has been unique, one important characteristic of major financial crises in the 1980s was an increasingly rapid market response to institutional weakness, especially when there were doubts about whether certain types of institutions would receive assistance or whether certain types of deposits or liabilities would be protected.
This experience suggests that while official safety nets have reduced the risks involved in placing retail deposits in financial institutions, there are still significant elements of market discipline imposed on such institutions. Stock markets, credit rating agencies, and large depositors and creditors are sources of such discipline. A gradual deterioration of a financial institution’s income and capital position typically results in a lower price for its equity, a reduced credit rating, a higher cost for credits and uninsured deposits, and a shortening of the maturities on the credits and deposits that it can obtain. If its financial position continues to worsen, a point can be reached at which large lenders and depositors simply stop extending funds to the institution. Moreover, a sudden, unexpected deterioration in a weak institution’s financial position can trigger an abrupt withdrawal of credits and a run of uninsured depositors.
This type of market discipline, however, is a two-edged sword as far as the stability of the financial system is concerned. On the one hand, such discipline creates incentives for owners and managers of financial institutions to undertake prompt corrective action when their institution’s financial position deteriorates; on the other hand, if managers ignore market signals or their financial position deteriorates suddenly owing to a macroeconomic shock, then a sudden withdrawal of funds could quickly lead to an institutional failure. An unmanaged failure could have systemic effects if it disrupts the payments system, contributes to the perception that other similar institutions are likely to face difficulty, or if the counterparty losses associated with its failure create difficulty for a broad range of other creditors, depositors, and (in the case of banks) correspondent banks.
In those cases where the authorities consider it necessary to assist a troubled institution (e.g., through arranged merger or receivership), the cost of such assistance is likely to be directly related to whether the authorities have the ability or the authority to close an institution or force it into receivership before it becomes economically insolvent (i.e., the market value of its net worth becomes zero or negative). The most costly episodes will occur when the authorities do not have the ability to act before the institution’s net worth is reduced to zero.73
The progressive integration of major financial markets, the expanding activities of foreign financial institutions in major domestic markets, and the growing responsiveness of large depositors to institutional weakness have raised new issues regarding the provision of emergency liquidity assistance. The events of October 1987 and August 1990 demonstrated the speed with which large financial and macroeconomic shocks can spread across major domestic and offshore financial systems. This suggests that during a major crisis there could be a sharp simultaneous increase in the demand for liquidity across all major markets, or that the demand for short-term liquidity in a particular currency (e.g., the U.S. dollar) could emerge in offshore markets even when the domestic markets of the currency are normally closed for business. The growing importance of the branches and subsidiaries of foreign banks (and other financial institutions) in major domestic markets also raises the issue of whether these institutions should have access to the domestic central bank’s discount window during a liquidity crisis.
As mentioned, a fundamental problem with regard to risk taking is that an official safety net may make depositors and possibly creditors indifferent to the risks assumed by financial institutions, except in unusual circumstances. If depositors and creditors do not adequately monitor the activities of a financial institution and demand that appropriate risk premiums be incorporated in the interest rates they receive, then the incentives of the institution’s managers and owners to control the risks they face could be blunted. In this situation, the main constraint on risk taking may be government capital requirements, regulation, and supervision rather than market discipline. Although elements of market discipline remain in all major financial systems, a key issue in some countries, such as the United States, is whether those elements and prudential supervision are sufficient to provide adequate constraints on risk taking, especially by managers of institutions that are nearly insolvent.
This moral hazard problem can affect both the distribution of credit within a given financial structure and the evolution of that financial structure over time. The underpricing of risk that can arise because of the safety net can artificially expand the resources allocated to finance risky projects. This may well crowd out investments with lower risk but also somewhat lower rates of return. When this occurs, the effect may be to shift the overall mix of investment in the economy toward high-risk and possibly high-return projects. Any shift toward greater funding of high-risk projects can require increases in the effort and resources used by supervisors and regulators to monitor and modify behavior. Institutions that are subject to such enhanced supervision also incur a variety of reporting and regulatory costs that reduce the implicit subsidy they receive from the safety net. These costs can in turn stimulate efforts by regulated institutions to shift some activities to either affiliated institutions or offshore markets. Associated changes in corporate structure or location may allow an institution to escape some regulatory or supervisory costs, while at the same time its domestic or foreign affiliate could be viewed by market participants as benefiting from implicit protection by the domestic official safety net.74 If these changes were to occur, there could be an artificial stimulus provided to activities in offshore markets and the scale of gross capital flows.
An official safety net can also influence the evolution of an economy’s financial structure. A key issue is whether the safety net effectively provides a net subsidy to some segment of the financial sector. Depending on its design, an official safety net can generate diverse benefits and costs for different financial institutions. For example, the availability of official deposit insurance (especially when it is underpriced), and the perception that short-term emergency liquidity assistance is likely to be available to an institution during a crisis period, can reduce its funding costs. Deposit insurance can reduce the risk premium that depositors demand be incorporated into the interest rate they receive on their deposits. Moreover, to the extent that creditors and owners perceive that the presence of an official safety net will enable an institution to better service its debts and equity claims over time, then the value of this assistance will also be reflected in the market prices of the institution’s securities (both debt and equity). In addition, depository institutions may receive an indirect benefit if the central bank’s operations in the payments system effectively reduce the costs of making payments and securities transfers.75
Partly offsetting these potential benefits are a variety of costs that are incurred by institutions that are protected by an official safety net. As already noted, depository institutions are typically faced with minimum reserve requirements on deposits, deposit insurance premiums, minimum capital requirements, and the costs of periodic reports to supervisory authorities. In addition, these institutions may face restrictions on the activities they can undertake (such as the separation of commercial and investment banking in Japan and the United States).
Although there is considerable debate about the scale and distribution of the net benefits or costs associated with the existence of an official safety net, there is no doubt that it can sharply alter the incentives for owning and operating different types of bank and nonbank financial institutions. For example, a net subsidy to a particular class of institutions can over time raise the proportion of total financial services provided by financial institutions that are protected by the safety net. Nonetheless, the value of the net subsidy received by different institutions or groups of institutions can evolve in unanticipated ways not necessarily connected with changes in the structure of the official safety net itself. Other regulatory changes and advances in telecommunications and computer technologies can affect the value of the subsidy and alter the competitive position of different types of financial institutions.76
Proposals for Reform of the U.S. Official Safety Net
The insolvency of a large number of thrift institutions and growing evidence of weakness in the income and balance sheet positions of commercial banks in the United States have raised concerns about the viability of the government’s deposit insurance funds and the adequacy of supervision and regulation of depository institutions. (Key components of the U.S. official safety net are given in Box 1.) These concerns led the U.S. Congress to call for a comprehensive study of the deposit insurance system by the U.S. Treasury.77 The study, entitled “Modernizing the Financial System: Recommendations for Safer, More Competitive Banks” was issued in February 1991; its major recommendations are summarized in Box 2. This section examines the proposals that were put forth during public debate while the treasury study was being prepared; they concern reform of not only the deposit insurance system per se but also the overall structures of the official safety net and the financial system. To set the stage, the principal factors contributing to the current financial weakness in the U.S. financial system are examined briefly, and this is followed by a discussion of the reforms that were proposed to strengthen the financial system and improve the cost-effectiveness of the official safety net.
Box 1Key Components of Safety Net
Lender of Last Resort
The Federal Reserve has broad powers to provide liquidity to banks and, in exceptional circumstances, to extend liquidity support to other institutions. Under the Federal Reserve Act, any Federal Reserve bank may make advances to any member bank provided they are secured to the satisfaction of the Federal Reserve bank. The Federal Reserve may also discount eligible trade bills; for ease of operation almost all discount window borrowings are in the form of advances. The International Banking Act of 1978 enables the Federal Reserve also to make advances to any branch or subsidiary of a foreign bank that holds reserves with the Federal Reserve; the Monetary Control Act of 1980 extends reserve requirements to non-member banks and provides for their access to the discount window on the same basis as member institutions. In addition, individuals, partnerships, and corporations other than depository institutions may be granted access to the discount window in unusual and exigent circumstances; no direct lending to such borrowers has been undertaken since the 1930s.
The Federal Reserve’s Regulation A establishes the terms under which credit may be extended to eligible banks and other institutions. The terms that pertain to the lender-of-last-resort function of the Federal Reserve are primarily those of extended credit advanced for other than seasonal purposes. Other extended credit is available to any eligible depository institution that is experiencing liquidity strains owing largely to exceptional circumstances affecting that institution alone. If the credit is outstanding for more than sixty days, the rate on extended borrowing is raised above the basic discount rate according to a schedule. The rate that would be applicable to loans made to individuals, partnerships, and corporations in unusual and exigent circumstances would be established at the time of the loan.
The Federal Reserve may provide extended credit either to support an explicit program of management and financial reform, with a view to returning a troubled depository institution to a sound basis, or to make available bridge financing while arrangements are completed for merger with a stronger institution or for an orderly closing. In such cases, the Federal Reserve works in conjunction with the appropriate federal and state supervisory agencies and the Federal Deposit Insurance Corporation (FDIC).
The FDIC provides explicit guarantees of deposits and assists in the resolution of failed institutions.1 The FDIC, an independent government agency established in 1933, operates two separate deposit insurance funds, the Bank Insurance Fund for commercial banks and some savings banks and the Savings Association Insurance Fund (SAIF) for thrift institutions.2 All member banks of the Federal Reserve System are statutorily required to carry federal deposit insurance, as are most nonmember state-chartered banks. Similarly, all federally chartered thrift institutions are required to join SAIF; participation is optional for some state-chartered institutions. The FDIC’s explicit insurance coverage is limited to deposits up to a ceiling of $100,000 a deposit, including interbank deposits and those in foreign currency. Under the International Banking Act of 1978, U.S. branches of foreign banks are eligible for FDIC membership; such branches must pledge to the FDIC a surety bond to secure the payment of domestic deposits. The FDIC charges a premium of 0.195 percent of insured, domestic deposits for its guarantee.3
The FDIC may provide assistance to depositors in a number of ways. First, it may be appointed as receiver of a failed bank or thrift institution by its chartering agency. In this case, it pays depositors their net balances up to the insurance limit and proceeds to liquidate the bank’s assets. Uninsured depositors and other general creditors of the bank generally do not receive either immediate or full reimbursement of their claims. Second, it may undertake a so-called purchase-and-assumption transaction. Under this approach, a buyer purchases all or some of the failed bank’s assets and assumes its deposit liabilities and certain of its liabilities. An important difference between a payoff and purchase-and-assumption transaction is that in the latter case all depositors, both insured and uninsured, receive full payment on their claims. General creditors also normally receive full payment on their claims. Third, it may transfer insured deposit and secured liabilities to another institution; uninsured and unsecured liabilities remain in receivership. Technically, the receiving bank is not purchasing the failed bank but rather acting as an agent for the FDIC by assuming the insured deposits. The FDIC pays the accepting institution an amount equal to the liabilities less any purchase price. Fourth, it may provide direct financial assistance to a bank that is in danger of failing. Generally, the FDIC’s assistance covers the difference between the market value of the bank’s assets and liabilities; new capital is injected by private investors. Finally, the FDIC may own and operate temporarily a bank until a permanent solution can be arranged.
Box 2U.S. Treasury Recommendations
On February 5, 1991, the U.S. Treasury Department issued its report on the federal deposit insurance system, entitled “Modernizing the Financial System: Recommendations for Safer, More Competitive Banks,” as required by the Financial Institutions Reform, Recovery, and Enforcement Act of 1989. In general, the recommendations aim to increase the market discipline faced by banks, to restructure banking regulation, and to eliminate many of the existing geographic and functional restrictions on banks.
Recommendations to increase the market discipline faced by banks include a phased reduction in the scope of deposit insurance coverage and the introduction of deposit insurance premiums related to the risks inherent in an institution’s lending activities. Explicit coverage of deposit insurance would include only $100,000 a depositor at each institution for ordinary accounts and $100,000 a depositor at each institution for retirement accounts, eliminating the use of joint accounts and trust accounts to expand insurance coverage.1 Explicit insurance coverage would exclude certain fiduciary deposits, those of professionally managed pension funds, and brokered deposits. To curb the implicit scope of deposit insurance, the Federal Deposit Insurance Corporation (FDIC) would consistently use an existing legal provision that enables it to expose nondeposit creditors to normal pro-rata bankruptcy losses even if uninsured deposits are made whole by the FDIC’s actions. Also, proposed legislation would require the FDIC to use the least costly resolution method to limit the coverage of uninsured deposits and other bank liabilities, although the U.S. Treasury and Federal Reserve could decide to protect uninsured deposits when “systemic risks” are judged to be present. Two recommendations are made regarding risk-based deposit insurance premiums. The first, for the short-term, would authorize the FDIC to establish risk-based premiums as a private insurer would, with capital levels used as the principal measure of bank riskiness. The second, for the longer-term, would establish a demonstration project to introduce a private insurance market into the process for pricing bank insurance premiums.
The main recommendation to restructure banking supervision is a system of capital-based regulation that rewards a well-capitalized bank with scope for expanded activities and expedited regulatory procedures and that subjects an undercapitalized bank to prompt and increasingly strong corrective actions. Measuring bank capital would be improved through annual on-site examinations for all banks, more accurate provisioning for loan losses, and increased market value reporting of assets and liabilities. Interest rate risk also would be included in risk-based capital standards. For undercapitalized banks, supervisory actions could include dividend restrictions, growth constraints, limits on risky activities of the bank and its affiliates, and management dismissals; such actions would intensify as a bank’s capital declines. In addition, to minimize losses that could be incurred by the deposit insurance fund, regulators would be provided with new legal authority or other means to close a failing bank promptly, while it still has some low level of positive book capital.
With regard to regulatory agency restructuring, the report proposes to consolidate responsibilities for banking regulation in the treasury and the Federal Reserve. Within the treasury, the Office of the Comptroller of the Currency and the Office of Thrift Supervision would be consolidated into a new Federal Banking Agency (FBA) that would supervise all national banks, their bank-holding companies, and all thrift institutions. The Federal Reserve would supervise all state-chartered banks and their holding companies, while the FDIC would focus solely on the administration of the deposit insurance system and the resolution of failed banks and thrifts.
Also recommended is reform of the geographic and functional restrictions on banks to improve their competitiveness and attract capital to the banking industry. Specific recommendations include (1) immediate authorization for a national bank to branch into any state in which the bank’s holding company is permitted to acquire a bank, and (2) authorization, subject to a three-year delay, for a bank-holding company based in one state to own a bank in any other state. The recommendations would thus permit full interstate branching after the three-year hiatus, effectively ending the branching restrictions of the McFadden Act of 1927. With regard to restrictions on functional activities, the proposed system of capital-based regulation would permit well-capitalized banks to affiliate with securities firms, mutual funds, and insurance companies within a new financial services holding company structure. The long-standing separation of banking from securities brokerage and underwriting activities under the Glass-Steagall Act would cease; while a proposed authorization for commercial firms to be permitted to own financial services holding companies would erode the separation of banking and commerce.
To prevent an expansion of the federal safety net to nonbanking financial and commercial activities, a set of safeguard measures is proposed. They would include the authority for only a well-capitalized bank to affiliate with other types of financial firms; functional regulation of the bank and its affiliates; restrictions on a bank’s lending to its affiliates and holding company; limits on a bank’s dividend payments when capital is low; regulatory notification of large transfers between a bank and an affiliate; and authority for the regulator to limit bank lending to its securities affiliate or certain customers of the affiliate.1 The current explicit coverage of individual deposit insurance is $100,000 a deposit, with no firm limit on the number of such deposits per individual. Under this system, an individual may open numerous separately insured accounts within a single institution through separate insured “capacities” (e.g., joint accounts, trust accounts, individual retirement accounts, and so on).
A key factor stimulating discussions of reforming the U.S. financial system has been the emergence of widespread difficulties in the savings and loan industry. The origins of the savings and loan crisis can be traced to the late 1970s and early 1980s and illustrate the important role of capital adequacy requirements and prudential supervision in promoting a cost-effective safety net. Because of tax incentives and regulations, savings and loan institutions ended the 1970s holding long-term, fixed interest rate assets (primarily residential mortgages) funded by savings and time deposits with shorter maturities. This maturity mismatch exposed thrifts to considerable interest rate risk, which was realized when short-term interest rates rose relative to the fixed return on their assets as inflation accelerated in the late 1970s and was then followed by a sharp monetary tightening in the early 1980s. By the time that greater interest rate stability had returned, the thrift industry had incurred such extensive losses that it was left with virtually no capital. Authorities responded to these developments by relaxing the capital adequacy standards for savings and loans, and, at the same time, by expanding the lending activities that they were allowed to undertake, the deposit-taking powers, and the coverage of deposit insurance. These actions, in effect, permitted insolvent or significantly undercapitalized institutions to continue their independent operation presumably in the expectation that the savings and loans would use their new freedom to undertake activities that would allow them to earn profits and rebuild their capital positions.
Shortfalls in prudential supervision enabled many of these institutions to undertake relatively risky investments that were often funded by offering higher-than-average interest rates on government insured deposits.78 Lower energy and agricultural prices and regional downturns in real estate prices in the second half of the 1980s led to a sharp increase in defaults in loan repayments by borrowers from many savings and loan institutions, which, in turn, caused a growing number of insolvencies of these institutions. In consequence, the savings and loan deposit insurance fund exhausted its resources, and a major injection of funds from the federal government was required.
While the most serious problems in the thrift industry were being addressed, a deepened concern arose about the financial strength of U.S. depository institutions, particularly commercial banks. Throughout the late 1970s and 1980s, the U.S. banking industry faced increased competition from other domestic providers of financial services. For example, shares in money market mutual funds increased to 7¾ percent of household holdings of credit market instruments at the end of 1989 from ½ of 1 percent at the end of 1978. At the same time, commercial paper increased from 2 percent of nonfinancial corporations’ total credit market liabilities to 5 percent. These competitive developments reflected in part recent advances in information processing and communications technologies that facilitated the operation of money market mutual funds and eliminated much of the traditional advantage banks possessed in assessing the credit risk of prime corporate borrowers.
As the banking industry faced increased domestic competition, its profitability weakened. While most banks continued to be profitable, the annual rate of return on equity of all insured commercial banks declined from an average of 12¾ A percent in 1975–81 to an average of 9¾ A percent in 1982–89. Much of the decline in profitability was concentrated in money center banks. Their annual rate of return on equity fell from an average of 13¼ percent in the earlier period to an average of 6¾ percent in the later period.
U.S. financial institutions also faced increasing competition from foreign financial institutions both within the United States and in offshore and major domestic markets in other countries. Prospective developments in the EC and in Japan also suggested to many observers79 that some of the existing restrictions on the activities of U.S. financial institutions may make it difficult for those institutions to maintain or improve their competitive position in global financial markets. In particular, the movement of the EC toward community-wide banking and the Japanese discussions of reforming the Article 65 separation of commercial and investment banking were regarded as implying that institutions from these regions will soon have access to a larger deposit base and will be able to offer a broader range of financial services than U.S. institutions.
As noted in Section III, the provisions of the Glass-Steagall Act of 1933, which required the separation of most commercial and investment banking activities, and the McFadden Act of 1927, which limited bank branching across state lines, have been cited as playing a key role in limiting the services that can be offered by U.S. commercial banks and the development of a national deposit base. While U.S. banks have to a degree escaped some of these limitations through the operations of their subsidiaries in financial markets in offshore centers and other countries, they may still lose some of the economies of scale and scope associated with integrated domestic and international operations and a broader national deposit base. Moreover, restrictions in bank ownership may inhibit bank restructurings and thereby limit the extent to which the banking industry responds to changes in its competitive environment.
These considerations contributed to the view, advanced also in the U.S. Treasury study, that the reform of the U.S. financial system must extend beyond changes in the deposit insurance system and should encompass more comprehensive changes in capital adequacy standards, prudential supervision, and geographic and activity restrictions on financial firms. The structural issues that have been regarded as important include (1) the scope and pricing of deposit insurance guarantees; (2) the appropriate minimum capital adequacy standards; (3) how prudential supervision should be improved; (4) the appropriate relationship between banking and other financial services, and between banking and commerce; (5) the nature and extent of firewalls between parent and subsidiary institutions; and (6) greater use of market value accounting in evaluating the positions of financial institutions. In making recommendations for changes in the financial system, the authorities have indicated that they will be guided by the need to improve the competitiveness of U.S. financial institutions and markets as a means to strengthen their capitalization, to maintain the stability of the financial system, and to limit the exposure of the taxpayer through the official safety net.
Awareness of the potential costs that can be created by the moral hazard problem associated with the safety net in general and deposit insurance in particular has stimulated a discussion of how best to develop “…an institutional framework that, to the extent possible induces banks both to hold more capital and to be managed as if there were no safety net, while at the same time shielding unsophisticated depositors and minimizing disruptions to credit and payment flows.”80 The reform proposals advanced in various quarters, while the study was being prepared, included establishing 100 percent reserve (or narrow) banking, introducing risk-based deposit insurance premiums, implementing depositor or other private co-insurance, raising capital requirements, establishing prompt closure rules for troubled institutions, and imposing restrictions on banks’ asset choices.
One proposal was to establish “narrow” banks that would invest only in high-quality, short-maturity, liquid investments and recover their costs for providing checking accounts and wire transfers from user fees. Moral hazard would be regulated away because these banks would not be able to invest in risky assets, and instability would be over-come because the deposits would be fully backed by safe and liquid assets. Establishing such narrow banks would obviously require a drastic change in the current U.S. financial structure and that of other industrial countries. Moreover, banking organizations would presumably locate their business and household credit operations in nonbank affiliates funded by uninsured deposits and securities issued on capital markets. Unless these organizations relied primarily on long-term funding, they would be subject to the same risks of creditor and depositor runs that any uninsured institution would normally face. If such a run were to occur, then the authorities would be confronted with the issue of either assisting these institutions (which again raises the moral hazard issue) or allowing the credit system to be severely disrupted.
Another proposal, which would not require a fundamental restructuring of major financial systems, would be the use of risk-based deposit insurance premiums.81 The basic idea is to make the price of deposit insurance reflect the riskiness of the bank’s portfolio, thereby reducing the subsidy for risk taking and achieving a more equitable distribution of the costs of deposit insurance across banks. To effectively administer such risk-based premiums, authorities must have some way of accurately assessing risk, which would require either the collection and analysis of extensive information about insured institutions or the reliance on market information to assess risk (e.g., through the use of reinsurance by the FDIC). Either approach has limitations in terms of cost and accuracy, and even if risks can be properly evaluated, the range of premiums necessary to induce significant behavioral changes in portfolio management might have to be many multiples of the existing premium. Nonetheless, while no major industrial country currently prices government deposit guarantees on a risk-adjusted basis, some broad-risk categories could be established and priced.82
Co-insurance makes the insured bear some of the cost of a bad outcome and has been widely used in the provision of private medical, fire, and automobile insurance. It has been argued that, if depositors were insured only up to, say, 90 percent of their deposits, they would have an incentive to monitor their banks’ activities.83 Co-insurance could also be designed to protect small depositors by fully insuring deposits up to a certain size and then providing only partial protection for deposits in excess of that amount. Just as with current arrangements, such coinsurance would only have its desired effects if it was credibly established that large depositors would experience losses when a bank failed. And, if such credibility was attained, there would then be strong incentives for large depositors to run at the first sign of trouble. Indeed, even under current arrangements, where large depositors have generally not suffered losses, runs of uninsured deposits from banks under stress have occurred.
As already noted, another approach to reducing the moral hazard problem has involved the implementation of risk-based capital adequacy requirements, which reflect both portfolio and off-balance sheet risk. In addition, it has been proposed that these capital adequacy requirements could be strengthened by prompt corrective action whenever a bank’s capital position is not consistent with minimum capital standards. This could initially involve forcing the bank to make credible commitments to raise additional capital, which, if not promptly met, would lead the authorities to require dividend reductions, restrictions on asset growth, limitations on the use of insured deposits, and sales of assets or affiliates with the resources used to rebuild capital. If these actions were not sufficient, then forced mergers, divestitures, and, if necessary, a conservatorship could be required while the bank still had some positive equity capital.84 In essence, the combination of enhanced capital adequacy requirements and prompt corrective action would be designed to simulate market pressures from risk taking without creating market disruptions. Nevertheless, even if these policies strengthened bank capital positions and the quality of their assets, they would not eliminate bank failures; their proponents have argued that some form of deposit insurance would still be necessary.
While there is wide support for strengthening the capital positions of financial institutions, concerns have been expressed that implementing these requirements when the income and capital positions of financial institutions are weak could adversely affect the availability of credit. This reflects the view that these institutions may have few options other than to reduce the level or growth of loans.85 Such reduced lending could adversely affect economic activity if otherwise creditworthy firms could not obtain funding for their current operations and investments. Proponents of enhanced capital adequacy have argued, however, that the activities that are likely to be curbed are exactly those high-risk activities that contribute the most to the moral hazard problem. Moreover, creditworthy enterprises have a variety of sources of funding for their activities; and financial institutions that are better capitalized are likely to be better able to adapt to technological change and macroeconomic shocks.86
Although all major industrial countries have acted to enhance capital adequacy standards and improve prudential supervision, concern has been expressed that these steps may not be adequate to overcome the moral hazard problem, particularly for banks undertaking a wide range of activities. As a result, some observers have suggested that new restrictions on the banks’ or their affiliates’ activities may be needed.87 The concern appears to reflect the view that it is becoming increasingly difficult, in a world of rapidly changing telecommunication and computer technologies and financial innovation, for the supervisory authorities to adequately evaluate the risks inherent in banks’ on- and off-balance sheet activities. This situation arises not only because of the difficulty involved in gathering timely information about the domestic and international activities of a bank and its affiliates but also because it is inherently difficult to evaluate the risks in new financial products. As a result, it has been argued that some new restrictions may have to be placed on the activities of deposit-taking institutions that are protected by the safety net or to force protected institutions to undertake certain activities in separately capitalized affiliates.
In most major financial systems, some restrictions on the activities of financial institutions still exist despite the extensive liberalization that has taken place over the past two decades.88 As already noted, for example, there is still a separation of commercial and investment banking in Japan and the United States, and, in a number of major industrial countries certain types of activity must be carried out in separately capitalized affiliates. An objective of these restrictions has been to try to ensure that the failure of institutions not protected by the safety net would not lead to problems of other (perhaps affiliated) institutions that are protected. Such restrictions have become increasingly difficult to enforce and may entail significant costs. Commercial banks from Japan and the United States regularly carry out investment banking activities in the Eurocurrency markets, and recent experience suggests that it may be difficult to maintain a firewall between a financial institution and its affiliates during a crisis.89 For example, when Drexel Burnham Lambert failed to meet its debt-servicing obligations in early 1990, the separately capitalized government securities affiliate, which had not engaged in unusually risky activities, found that it lacked access to credit.90 Activity restrictions may also weaken the profitability of banks by restricting their access to profitable new markets for financial services and thereby contribute to an erosion of their capital. In addition, restrictions on bank’s activities, location, and ownership may have weakened the industry’s response to changes in its competitive environment.91 Indeed, recent calls for liberalizing the regulations governing bank ownership, and those contained in the Glass-Steagall and McFadden Acts, aim at restructuring the banking industry to strengthen its competitiveness.
Payments, clearance, and settlements systems encompass the transfer of both money and securities. In some cases, this involves the transfer of money as the counterpart to the purchase of goods; whereas, in other cases, there may be an exchange of money for securities. In either case, these systems complete the transfer of money and securities and record the change in ownership.
Some systems accumulate payment instructions over a period of time and only net debits and credits are entered into the settlement accounts of the members at the end of the settlement period. In such a payments system, there is the credit risk that a member would not have sufficient funds to meet its net debit at the time of settlement. In other systems, such as Fed-wire in the United States, the central bank guarantees that payments instructions will be fulfilled (payment finality) even if a financial institution fails. A detailed analysis of systemic risk in payments systems is undertaken in Folkerts-Landau (1990b).
For example, Heller (1989) has proposed that the U.S. Federal Reserve could intervene in the stock index futures market to help stabilize equity prices in a crisis period.
If investment decisions are not reversible, the uncertainty created by a financial crisis may depress business fixed investment and household spending on durables. For example, a firm that makes an investment decision must trade off the extra return from early investment against the benefits of increased information gained by waiting. In a financial crisis, the firm would be less sure about the return it will earn on an investment, and the benefits gained by waiting for more information would increase and investment would decline. For theories of irreversible investment decisions under uncertainty, see Cukierman (1980) and Bernanke (1983), and, for empirical evidence of their significance, Romer (1990).
When providing emergency liquidity assistance, central banks typically limit their credit risks by acquiring the most liquid and highest quality assets of the banking system. This reflects the fact that open market operations usually involve the purchase of government securities; whereas the assets that banks discount at the central bank must usually meet high standards for both liquidity and quality.
For example, both private and official assistance were given to Continental Illinois in the United States in 1984.
In Japan, Yamaichi Securities was assisted in 1965, and in Canada, several modest-sized commercial banks in the western provinces received substantial assistance during the 1980s.
The rationale for this vagueness was recently stated by Corrigan (1990, p. 14) as:
With any troubled institution, but especially in the case of large institutions, I believe that the workings of both the safety net and market discipline will be better served in a context in which the authorities maintain a policy of what I like to call “constructive ambiguity” as to what they will do, how they will do it, and when they will do it…. The circumstances associated with a particular case, the setting in which it occurs, and the assessment of the relative costs and benefits of alternative courses of action will always have to be looked at case by case. But in no case should it be prudent for market participants to take for granted what actions the authorities will take and certainly in no case should owners and managers of troubled institutions—large or small—conclude that they will be protected from loss or failure.
In systems with securities houses and brokerage firms, as well as banks, various investor protection arrangements have been created to help ensure that the securities and funds of investors are protected if a securities house or brokerage firm should fail.
The capital controls used by Japan and many European countries in this period also limited external risk-taking activities.
For example, when Bankhaus Herstatt failed in 1974, the German banking association set up an emergency fund to reimburse small depositors, but some larger depositors reportedly lost up to 55 percent of their funds.
A comparison of deposit insurance arrangements can be found in U.S. Congressional Budget Office (1990).
Despite these legal limits on deposit insurance, the authorities in all major countries have at times taken steps to protect all depositors from loss. As a result, the number of instances in which either small or large depositors have incurred actual losses has been rare.
Even when there are limits on the extent of deposit insurance, the authorities face a “time inconsistency” dilemma (a policy that works best in the short run may not be best in the long run and vice-versa) with regard to uninsured large deposits. Once a bank is on the verge of failing, it often appears that the best policy is to protect all depositors, both insured and uninsured, so as to prevent a disruption of the financial markets, especially if the bank is large. Uninsured depositors will then learn over time that, whatever the announced policy, their deposits are protected. In this situation, uninsured depositors will also have a reduced incentive to monitor the activities of banks, and an element of market discipline will be lost.
The incentive to substitute deposits for equity as a source of funding would be increased if deposit insurance is provided at a low cost to financial institutions.
Until the 1870s, the average ratio of capital to assets exceeded 30 percent.
Capital adequacy standards for securities houses are also being discussed by the International Organization of Securities Commissions (IOSCO). While progress has been made, there has been a problem in defining equivalent capital standards across systems where banks’ participation in security activities and transactions differs markedly (e.g., in Japan and the United States versus Germany).
Even prior to these liberalizations, some of these restrictions were evaded through the emergence of new financial products and institutions and operations in offshore financial markets.
The restrictions on portfolio holdings encompassed limits on bank lending to individual borrowers, on the concentration of loans to a single industry or type of borrower, and on the types of business that could receive loans (e.g., banks in some countries may not make residential mortgage loans). Other financial institutions, such as insurance companies and pension funds, were required to invest in assets of a certain minimum credit rating and were often limited in terms of the foreign assets they could hold.
A large private offshore payments system is the Chase Manhattan U.S. dollar payments system, which operates in Tokyo when U.S. domestic payments systems are closed.
The caps were defined as a multiple of the depository institution’s capital. These caps were subsequently extended to CHIPS. A more detailed discussion of this risk-management policy can be found in Folkerts-Landau (1990b).
The fee was to equal 25 basis points on the daily average overdrafts, less a deductible equal to 10 percent of capital, and it was to be phased in over a three-year period beginning in 1991.
The failure of Drexel Burnham Lambert disrupted the settlements system for mortgage-backed securities in the United States.
The Bank for International Settlements published in November 1990 a “Report on Netting Schemes,” prepared by the Committee on Interbank Netting Schemes composed of senior officials from the Group of Ten central banks, that examined how netting arrangements could help reduce the scale of transactions and risks in clearance and payments systems for foreign exchange and securities.
A survey by the Bank for International Settlements indicated that the average daily foreign exchange trading volume in the major markets had reached $640 billion in April 1989.
It is generally thought that these settlement risks are greatest for transactions involving trade of Japanese yen for U.S. dollars in Tokyo. A party that sells yen in exchange for dollars must irrevocably pay out the yen at least eight hours and most often fourteen hours before it receives payment in U.S. dollars.
This would leave each participant with net amounts due to or due from the clearing house equal to its multilateral net positions vis-à-vis other participants in the system. The clearing house would become the counterparty for each contract submitted by a pair of participants, and the clearing house would keep a running, legally binding net position for each participant for each currency and each date.
Schwartz (1988) also argued that, between 1930 and 1933, there were three system-wide banking panics in the United States and two in Germany. However, she characterized the events that led up to the suspension of convertibility of the domestic currency into gold in Japan and the United Kingdom in 1931 as exchange crises rather than banking panics.
The costs to the U.S. authorities of rescuing Continental Illinois Bank were significantly lower (measured relative to the size of the bank’s assets) than those for the savings and loan industry (again measured relative to this industry’s total assets). In part, this reflected the fact that a run of large depositors forced the authorities to intervene in the Continental Illinois case when the bank still had a modest positive net worth. In contrast, many savings and loans that ultimately failed were allowed to operate with increasingly negative net worth for an extended period. These institutions could continue to operate in this condition only because they could rely on government insured deposits as a source of funding.
For example, this could occur if a bank shifted some of its deposit-taking activities to a foreign branch. Since these deposits might not formally be part of the deposit insurance program of the home country, they would not be assessed deposit insurance premiums. Nonetheless, if the authorities had previously taken steps during periods of institutional difficulty to protect offshore deposits, then there could be the perception that such actions would also be taken in the future.
For example, an institution’s payments costs could potentially be reduced if the central bank extends daylight overdrafts (at a low or zero cost) so that payments can be made even when the institution does not have sufficient funds on deposit at the central bank.
For example, Keeley (1990) argued that the deregulation of the financial system in the United States during the 1970s and 1980s, which removed some geographic restrictions on banks and allowed nonbank financial intermediaries to engage in activities previously restricted to banks, seriously eroded the market value (as reflected in the prices of bank equity) of a bank charter. Moreover, as the market value of bank charters declined, owners and managers of banks faced a smaller incentive to control the risks inherent in their loan portfolios because they had less to lose if they got into difficulty and their charter had to be surrendered.
This study was mandated by the Financial Institutions Reform, Recovery, and Enforcement Act of 1989.
The higher-than-average interest rates paid by these institutions indicate that a limited form of market discipline was being imposed.
For example, see the testimony by the Honorable Nicholas F. Brady, Secretary of the U.S. Treasury, before the U.S. Senate, Committee on Banking, Housing, and Urban Affairs, July 25, 1990.
Statement by A. Greenspan, Chairman of the Federal Reserve, before the U.S. Senate, Committee on Banking, Housing, and Urban Affairs, September 13, 1990, in Greenspan (1990), p.918.
These risk-based insurance premiums could also be combined with the use of risk-based capital requirements that will be discussed subsequently.
This type of approach has been used in establishing the new Bank for International Settlements capital adequacy standards, as noted above.
A related proposal has been to increase the use of subordinated debt as a source of bank capital. It has been argued that such nonrunable, but serially maturing, debt would provide for a periodic market evaluation of the bank and thereby introduce additional market discipline. Such subordinated debt could also serve as an additional buffer for the deposit insurance fund over and above the owners’ equity; however, the recent experience of some U.S. banks, whose credit ratings were downgraded, has shown that such subordinated debt can be quite expensive. In such situations, banks with weak financial positions may have only limited incentives to use such debt rather than relying on issuing additional equity or retained earnings. Nonetheless, institutions with strong income and capital positions may benefit from using subordinated debt to meet capital requirements.
Proponents have argued that this policy could be applied equally well to large and small banks.
However, Furlong and Keeley (1989) have argued that a profit-maximizing bank would add additional capital, rather than sell assets, in the face of higher capital requirements if deposit insurance implicitly subsidizes the cost of deposits to the bank.
In establishing higher capital adequacy requirements, supervisory authorities have also been confronted with the possibility that, if requirements were raised in one country in isolation, financial institutions would shift activities to affiliated institutions in other domestic or offshore markets (“regulatory arbitrage”). Alternatively, if higher capital requirements were imposed on one type of institutional structure (e.g., banks), financial conglomerates would shift activity to a different corporate structure (e.g., a leasing company). To avoid creating such incentives, supervisory authorities have implemented higher capital adequacy standards in the context of a multilateral agreement and have taken steps to improve the capital adequacy of a broad range of bank and securities market institutions.
The “narrow” bank proposal can be viewed as one such proposal, albeit an extreme version.
A more fundamental issue has been whether the U.S. banking structure should move from a system based on bank-holding companies to one based on universal banks either of the style employed in the United Kingdom or in Germany.
Even in normal periods, the protection of the safety net may be implicitly transferred to an affiliate. This can occur if a parent bank uses insured deposits to fund its loans to an affiliate or if the affiliate’s borrowing costs are reduced because the market perceives that difficulties in the affiliate would lead the authorities to support the affiliate in order to protect the parent.
Moreover, this affiliate also found it difficult to sell securities. This reflected the fact that, in the U.S. settlements system, the transfer of funds from the buyer to the seller of a security would take place within a shorter period than the transfer of title to the securities from the seller to the buyer. As a result, there was concern that the affiliate and the parent would enter into bankruptcy after the payment for the sale of a security by the affiliate had been received but before the title to the security had been transferred. The parent experienced similar difficulties in attempting to unwind its foreign exchange position and manage its positions in the commodities markets.